More than half of all Sun-like stars (spectral types F, G, and K) reside in binary or multiple star systems. Robust surveys show that binarity varies by spectral type (Lada 2006). M dwarfs represent the commonest spectral type in the Milky Way Galaxy, comprising at least 75% of all stars (Tarter et al. 2007). However, they occur preferentially in single-star systems, with only about 25% of all M dwarfs located in binary or multiple systems. Binarity increases among Sun-like stars of spectral classes K and G, with 57% of G stars residing in binary or multiple systems (Lada 2006).

Although comparable data are lacking on more massive stars, many investigators suggest that binarity continues to increase along with stellar mass through classes F through O, such that almost all of the most massive stars have binary companions (Lada 2006, Maiz-Apellaniz 2008). Nevertheless, these brighter spectral types represent increasingly smaller populations of stars, so their high proportions of binarity have little effect on the overall ratio of single stars. C.J. Lada concludes that about 66% of all stars throughout the Galaxy are single, and only one-third belong to binary or multiple systems (Lada 2006).

The relative proportions of single and multiple star systems will vary somewhat by location within the Galaxy. Surveys of the volume-complete sample of stars within 10 parsecs of the Solar System, conducted by the Research Consortium on Nearby Stars (RECONS), yield the following results: 69% of stars are single, 23% have a binary companion, and 8% occur in systems of three or more stars (RECONS 2008). Notably, the same survey finds that 72% of nearby stars are M dwarfs, similar to estimates for the entire Galaxy. (See the RECONS Web page.)

planet formation and stellar multiplicity

Although some earlier theoretical models placed strict limits on planet formation in multiple systems, precise radial velocity surveys have shown that exoplanets are found almost as readily in these systems as around single stars (Raghavan et al. 2006, Desidera & Barbieri 2007). In all cases planet formation is thought to proceed by planetesimal accretion within a primordial nebula (see Evolution of Planetary Systems).

Nevertheless, the dynamical configurations of binary and multiple systems imply evolutionary constraints that do not affect single stars. The most important such constraint is periastron – that is, the closest approach between members of a binary pair. Binary periastra can range from a small fraction of an AU (so close that in some systems the two stellar atmospheres are in direct contact, resulting in mass transfer from the smaller star to the larger) to as much as a few thousand AU (i.e., about 5% of a light year).

Desidera & Barbieri argue that planet formation may proceed virtually unimpeded around the members of a wide binary pair, which they define in terms of a periastron greater than 200 AU. They conclude, “The properties of exoplanets orbiting the components of wide binaries are compatible with those of planets orbiting single stars, except for a possibly greater abundance of planets on highly eccentric orbits” (Desidera & Barbieri 2007).

Binaries whose orbits bring them closer than 200 AU will probably experience some truncation of their protoplanetary disks, with smaller periastra corresponding to more extreme truncation. Nevertheless, the potential to form gas giant planets will probably remain in most systems whose disks extend substantially beyond their ice lines. For a Sun-like star this will be a radius of about 10 AU, corresponding to a likely binary separation of 50 AU (Pfahl & Muterspaugh 2006).

As an example, Rodriguez et al. (1998) describe a pair of newly-formed Sun-like stars in the L1551 molecular cloud in Taurus, each of which is surrounded by a compact disk of gas and dust. The two stars are separated by about 45 AU, and each circumstellar disk has a radius of about 10 AU. The researchers note the possibility of a larger, more tenuous envelope encompassing both stars, but offer no secure conclusions. The radii of the confirmed dust disks contrast strongly with those centered on such young single stars as Epsilon Eridani (radius 105 AU) and AU Microscopii (radius 120 AU). Despite their truncation, Rodriguez and colleagues conclude that each of the protoplanetary disks in L1551 is massive enough to yield a system of giant and terrestrial planets.

Binaries resembling L1551 may be the precursors of exoplanetary systems such as Gamma Cephei and Gliese 86, discussed below. In each of these systems, the binary components approach within 20 AU of each other, yet gas giant planets have nevertheless managed to achieve stable orbits.

Even closer binary pairs may still support systems of terrestrial planets, as demonstrated by numerical simulations. Quintana et al. (2007) found that binaries containing G, K, or M stars with a periastron of at least 10 AU could support accretion disks at least 2 AU in radius. Such disks had the potential to evolve into systems of two to five terrestrial planets within the same radius. Periastron separations smaller than 10 AU placed increasingly more severe limits on planet formation. Unfortunately, existing methods are incapable of detecting Earth-mass exoplanets.

More tightly bound binaries, with separations of a few AU or less, are generally excluded from radial velocity searches because they present special problems in data analysis. Nevertheless, simulations by Quintana & Lissauer (2006) indicate that a binary pair with semimajor axis smaller than 0.2 AU and eccentricity approaching 0 can support the evolution of a single planetary system very similar to our own. In such a case, an extensive protoplanetary disk would originally surround both stars, providing a far more fertile field for the evolution of planets than would a close binary pair like Gamma Cephei A and B. Unfortunately, current methods are unable to detect such systems, and in any case they may be rare. Close binaries that exceed the authors’ strict parameters – slightly wider separation, slightly stronger eccentricity – would be unlikely to retain any planets at all.

planets in nearby binary and multiple star systems

Planet-hosting binary and multiple systems may comprise stars of diverse spectral types at a wide range of orbital separations. A review of the 37 exoplanetary systems located within 20 parsecs yields 12 such systems. This means that about two-thirds of nearby exosystems center on a single star, similar to the overall percentage of single systems among field stars. Such a high proportion of binary or multiple hosts hints that single stars may be no more likely than binaries to harbor planetary systems.

So far, however, the sample of known planet-bearing multiples reveals a strong selection bias. The 12 nearby systems comprise fully one-third of the binary and multiple exoplanetary systems recently identified by Desidera & Barbieri (2007), meaning that the sample of host stars beyond 20 parsecs is strongly biased in favor of single-star systems. This bias implies not only that planets become more difficult to detect around more distant binaries, but also that binarity becomes more difficult to determine with increasing stellar distance. In fact, for many of the systems noted by Desidera & Barbieri, the host star’s binary companion was not identified until after the detection of its planetary companion encouraged more intensive observations of its immediate vicinity.

Among the 12 systems within 20 parsecs, 11 are binary and one is triple. The binaries can be further divided into three configurations according to mass and spectral type.

• The commonest configuration consists of one Sun-like star (spectral type F, G, or K) with a companion of much lower mass, either an M dwarf or a T dwarf. These parameters describe seven nearby systems: 54 Piscium, 55 Cancri, Upsilon Andromedae, Gamma Cephei, Tau Boötis, GJ 3021, and HD 189733. Within this group, 54 Piscium is notable as the host of a large brown dwarf (spectral class T); 55 Cancri and Upsilon Andromedae host multiple planetary systems and distant M dwarfs; Gamma Cephei is a subgiant or giant star with one gas giant planet and an M dwarf on a tight orbit; GJ 3021 hosts an eccentric gas giant and a very small M dwarf companion; Tau Boötis hosts a Hot Jupiter and a relatively large M dwarf on a highly eccentric orbit; and HD 189733 hosts a transiting Hot Jupiter and small, non-coplanar M dwarf.
  
  
• The second most common configuration includes one Sun-like star plus a white dwarf. Three such systems are located nearby: Gliese 86, a K1 star with a white dwarf at a semimajor axis of 18.4 AU; Epsilon Reticuli, a subgiant or giant of spectral type K2 with a white dwarf at 250 AU; and HD 147513, a G3 star with a white dwarf at 4450 AU. In each case, the main-sequence progenitor of the white dwarf was hotter and more massive than the known exoplanetary host star.
  
  
• The third configuration, consisting of two Sun-like stars, has only one instance within the Solar neighborhood. This is 83 Leonis, located at a distance of 18 parsecs (59 light years) in the constellation Leo. The primary is a G8 star of 1.01 MSOL (Desidera & Barbieri 2007) without any detected planets. The binary companion is a K2 star of 0.86 MSOL hosting a giant planet of sub-Saturn mass, traveling in an eccentric orbit with a semimajor axis of 0.12 AU. The current separation between the two stars is considerable, at about 515 AU (Desidera & Barbieri 2007).
  
  

The only exoplanetary system within 20 parsecs to include three stars is Gliese 777, located at a distance of 15.89 parsecs (52 light years) in the constellation Cygnus. The planetary host, a yellow star of spectral class G6, is accompanied by a pair of M dwarfs orbiting at about 3000 AU (Desidera & Barbieri 2007). This wide separation indicates that the M stars play no substantial role in the dynamics of the planetary system, which includes a Neptune-mass planet in a short-period orbit and gas giant at about 4 AU.

As noted above, binary and multiple systems can also be described in terms of their periastra. Among the 36 systems listed by Desidera & Barbieri (2007), at least 26 (72%) are separated by more than 200 AU. Such a preponderance of wide binaries among exoplanetary systems should not be taken at face value, however. It results from the fact that binaries with smaller periastra tend to be excluded from radial velocity searches, because their spectra cannot be distinguished reliably enough to assess the minute Doppler variations that signal the presence of an orbiting planet.

Among the more tightly bound exoplanetary systems (those with separations smaller than 200 AU), only three consist of binaries separated by less than 50 AU. These are Gamma Cephei, Gliese 86, and HD 41004 AB. These systems are especially important from a theoretical perspective, since the truncation of their protoplanetary disks must have had complex (and so far largely unknown) effects on the evolution of their planetary systems.

Several nearby binary and multiple systems, both with and without planets, are discussed in the pages linked below.

Last update April 2008